These are really cool renders and a good way to visualize vs how I've been doing it. I lost patience for using rendering and art modelling a long time ago.

What we did with the ScaleBirds concept and what I tend to think makes the most sense for any replica project that isn't being done by scaling original drawings via slide rule, is just find the size of airplane you want first. Then just adjust the scale of the original arbitrarily over top of the idealized template until it overlaps as close as possible. Then tweak from there either up or down to something that's vaguely round or nominal or fractional, or accept that the ideal final resulting machine will have so many subtle deviations from the original that you can call it an 80% when really it's the 81.42% fuse and a 82.14% wing with a 76.4% landing gear and a cowling that's stretched 106% athwartships and 98% longitudinally from exact proportion at 81.42% and the cockpit is at 93% scale with some canopy elements pretty-much unchanged from original and yet the whole thing is forward about 20% from where it should be. Ultimately if you're within 5% of a notional scale factor on 90% of the shapes, and stick the details in the relationally correct spots, no-one is going to be able to really say you're wrong.

I can't find any good data on one at this point, but would HIGHLY doubt a WAR 109 was truly 50%. Must be closer to 65-70% but everyone just knows that the WAR were 'generally' 1/2 scale with a very broad brush of what 1/2 means.

Then the real challenge is just whether your original 'ideal airplane' actually ends up what you wanted afterall. It can start getting into a real headscratcher if you think too hard about it. Maybe new engines come out that would affect the best possible design, or the detailed design stage reveals issues with fit that weren't fully appreciated during the initial layout stage. Maybe you get 87.6% thru the build and wonder why it wasn't just that 3% bigger so that a 34.79% more powerful engine could fit and allow the scale to be 130% more accurate by moving the cockpit 11.1% further aft, or maybe a 1.3% percentage increase of the wing volume would have allowed using a newly available nicer looking 81.31% landing gear scale vs the 76.4% you settled on due to sizing of available wheel/tire combos.

Fun times!

And then yes, what's interesting about the array of different warbirds at different scale levels is that you just got to pick from the different batches different warbirds and using the comparison of the person standing next to the nose, you can end up with a smattering of scales that end up with different warbirds at nearly the same final size and proportions. So a "5/8 P-40" can equate to roughly the "3/4 109" and so on.